How Smart Data Tech is Speeding Up Medical Research
Featured paper: Developing FHIR based Activity Libraries to Support Clinical Trial Direct Data Capture
Disclaimer: This content was generated by NotebookLM.
Imagine you are a scientist on the verge of discovering a cure for a rare disease. After years in the lab, it is finally time to test your medicine on real people in a clinical trial. This is the most crucial phase of medical research: a scientific study where doctors test new treatments to make sure they are safe and actually work.
But in the clinic, you hit a massive roadblock. Instead of focusing on patients, doctors and nurses are buried in paperwork. They write down vital signs and lab results, or type them into hospital computer charts. Then, they have to manually copy that exact same information into a completely different database for the clinical trial.
This process of copying and pasting is slow, expensive, and can easily lead to typos. Why are our medical systems still playing “telephone” with life-saving data?
Researchers Patrick Genyn and Andrew Richardson recently developed a solution. By using a smart coding standard called FHIR and graph technology, they built a way for medical systems to share data automatically, making clinical trials faster and safer.
The Great Translation Problem
To understand their solution, we first have to understand why medical databases do not talk to each other. It comes down to two major issues: structural interoperability and semantic interoperability.
These technical terms represent very simple concepts:
- Structural Interoperability is about the shape of the data. Imagine trying to plug a three-prong US power cord into a two-prong European outlet. They do not fit because they are shaped differently. In healthcare, different databases organize their fields in different structures.
- Semantic Interoperability is about the meaning of the data. Imagine one hospital writing “high blood pressure” in a patient’s chart, while another writes “hypertension.” A computer might think they are different conditions, even though they mean the same thing!
Today, hospital records and research forms are structured differently and use different vocabularies, meaning they cannot share data directly. Someone must manually copy data from hospital charts into research databases.
Enter FHIR: The USB Plug of Healthcare
To bridge this gap, the medical world is adopting a technology standard called Fast Healthcare Interoperability Resources, or FHIR (pronounced “fire”).
Think of FHIR as the USB plug of the healthcare industry. Just like a USB cable lets you plug almost any keyboard, mouse, or printer into any computer, FHIR allows different medical software programs to exchange information instantly and safely.
FHIR creates standardized, digital “building blocks” for things like patient details, vital signs, and lab tests. Because these blocks are identical across all FHIR-friendly systems, computers can share clinical records automatically.
Mapping the Journey with Graphs
Every clinical trial has a study plan called a Schedule of Activities (SoA). This is a timeline showing what measurements must be taken during each patient visit.
Genyn and Richardson used a math and computer science tool called directed graphs to map out these schedules.
Like a subway transit map where stations (nodes) are connected by tracks (edges), a clinical trial graph maps out patient visits and the specific medical tasks required at each stop.
Researchers can attach specific FHIR building blocks, like required laboratory tests, directly to these stations. Computers then read the map to know exactly what data to collect, without changing the schedule.
Standard Dictionaries: LOINC and SNOMED
To ensure computers understand measurements, researchers connected their graphs to medical “dictionaries” like LOINC and SNOMED CT.
- LOINC is a dictionary for lab tests and clinical measurements, assigning a unique code to every test (like bilirubin blood tests).
- SNOMED CT is a dictionary for physical procedures and clinical concepts, like drawing a blood sample.
By embedding these codes into the trial map, there is no confusion. No matter what terms a hospital uses, computers know exactly what the data means because of these universal codes.
Reusable Libraries: Lego-Style Science
Designing clinical trials from scratch wastes time. To fix this, researchers built activity libraries—collections of pre-coded medical tasks that work like Legos.
Instead of starting over for each study, researchers can grab pre-designed blocks from four specialized libraries:
- An observation library containing thousands of standard measurements, like heart rate.
- A specimen library for blood, urine, and other patient samples.
- An activity library for common clinical tasks.
- A synthetic data library for generating fake, realistic data.
Now, a research team can simply select pre-designed blocks from these libraries and snap them together to plan a study in hours instead of weeks.
Simulating the Future with “Synthetic” Patients
Before flying an airplane, pilots use a flight simulator. Similarly, researchers use “flight simulators” to test clinical trials before they begin.
Using their libraries, the researchers generated realistic synthetic data—highly accurate “fake” data that mimics real patient behavior, such as completing a visit or missing an appointment. By running this simulated data through their pipeline, research sponsors can test their computer networks and verify everything works before the trial starts.
Why This is a Game-Changer
This smart automation has huge real-world benefits:
- Less Admin Burden: Clinic staff can focus on patients instead of re-typing data into multiple systems.
- Fewer Mistakes: Eliminating manual entry reduces typos, making clinical trials safer and more trustworthy.
- Faster Cures: Automated data capture speeds up trials, bringing life-saving treatments to patients much sooner.
The U.S. Food and Drug Administration (FDA) is already evaluating using FHIR for clinical trials. By connecting hospital systems directly to research databases, we can build a faster, safer, and healthier future for everyone.